€4 ALASKAN GLACIER STUDIES the Kwik valley that now separate it from the Malaspina. These deposits are so thick that they would suffice to almost prevent further melting. Advance of Lucia Glacier in 1909. So far as we know, Lucia Glacier was receding continuously between 1890 and 1909. In 1906 there was some evidence, though not very conclusive, of a coming change, and an advance was predicted on the basis of an apparent increase in crevassing up the valley and on the assumption that a longer glacier, like the Lucia, might reasonably be expected to advance later than the shorter ice tongues, like the adjacent Atrevida.1 When we saw Lucia Glacier in July 1909 it was advancing as the Marvine and others were doing in 1906, but i^ad reached a less complete stage of advance than any of the 1906 glaciers. In view of the rapidity with which the advance passed down the glaciers, once it had started, as observed in 1906, it seems doubtful whether the crevassing observed in the upper Lucia Glacier in 1906 was in reality the forerunner of the 1909 advance. Had it been an indication of coming change one would expect that its effects would have been felt in the lower glacier by 1907. Still, our knowledge of the behavior of glaciers advancing under the impulse of snow supplied by earthquake shaking is too slight and too fragmentary to warrant definite exclusion of this suggestion of coming change. It may be true, for example, that only one or two of the tributaries had communicated the impulse of advance to the main glacier and that sufficient force had not yet been supplied to extend the crevassing farther down, while a year or two later the impulse from other branches, either more numerous or larger, gave sufficient impetus for a continuation of the thrust and for the extension of its effects down the glacier. This is an hypothesis at least worth retaining until the behavior of such advancing glaciers is better understood. Be this as it may, Lucia Glacier, when seen in July, 1909, almost exactly three years after our last observation of it in 1906, was utterly transformed, and was rapidly changing under our very eyes. Between Terrace Point and Lucia Nunatak, where we had so easily crossed in 1905 and 1906, the undulating moraine-covered surface was rent by a series of gashes (PL XVIII) which rendered the glacier impassable. These gashes were profound, flaring crevasses of great length, spaced a few yards apart, nearly parallel, and extending in the direction of flow of the glacier. The glacier had the appearance of being newly-crevassed by differential strain due to motion down the valley and toward the center of the glacier there was some indication of faulting, with the uplift on one side of the crevasses. The direction of the crevasses varied both below and above Terrace Point, but in each case changes in direction seemed attributable to change in direction of flow of the glacier. For example, near Lucia Nunatak the crevasses pointed toward the valley center, as if the ice flow here were deflected by the nunatak; and in the piedmont bulb portion the crevasses fanned out as the ice itself does to form the piedmont expansion. Between these parallel linear gashes there were very few, in fact scarcely any cross breaks, but there were long, undulating strips of serac, still covered with ablation moraine, over which travel would be easy; and it seemed probable that once one reached this gashed area it would be a matter of no great difficulty to cross the central part of the glacier by a zigzag course, threading one's way around the ends of crevasses. Along the margin, on both the east and west sides of the glacier, the condition of crevassing was widely different from that of the center. Here the ice was confusedly broken, »Tarr. R. S., Professional Paper 64, U. S. Geol. Survey, 1909, p. 80.